How happy is your pet? The Problem of subjectivity in the assessment of companion animal welfare

Abstract

The ability to evaluate the welfare of non-human animals accurately and objectively is influenced by a variety of factors including the nature of our relationships with them. Subjective biases in the perception of an animal’s quality of life can have either positive or negative consequences for its welfare and are likely to be particularly exaggerated in the case of companion animals, such as dogs, cats and other pet species, with which people tend to form strong anthropomorphic, attachment-based relationships. The consequences of these subjective biases are likely to be further exacerbated by the fact that many of the physical and behavioural attributes that humans find appealing, and have selected for, in companion animals, are inherently detrimental to their welfare. Using a range of examples, this paper explores some of the complex ways in which anthropomorphism and subjectivity can cloud our ability to make reliable judgements concerning the welfare of companion animals, even in the face of seemingly obvious and overt indicators of pain and suffering. © 2019 Universities Federation for Animal Welfare The Old School, Brewhouse Hill, Wheathampstead, Hertfordshire AL4 8AN, UK.

Full text

57
© 2019 Universities Federation for Animal Welfare
The Old School, Brewhouse Hill, Wheathampstead,
Hertfordshire AL4 8AN, UK
www.ufaw.org.uk
Animal Welfare 2019, 28: 57-66
ISSN 0962-7286
doi: 10.7120/09627286.28.1.057
How happy is your pet? The problem of subjectivity in the assessment of
companion animal welfare
JA Serpell
School of Veterinary Medicine, University of Pennsylvania, Philadelphia PA 19104, USA; email: serpell@vet.upenn.edu
Abstract
The ability to evaluate the welfare of non-human animals accurately and objectively is influenced by a variety of factors including the
nature of our relationships with them. Subjective biases in the perception of an animal’s quality of life can have either positive or
negative consequences for its welfare and are likely to be particularly exaggerated in the case of companion animals, such as dogs,
cats and other pet species, with which people tend to form strong anthropomorphic, attachment-based relationships. The conse-
quences of these subjective biases are likely to be further exacerbated by the fact that many of the physical and behavioural attrib-
utes that humans find appealing, and have selected for, in companion animals, are inherently detrimental to their welfare. Using a
range of examples, this paper explores some of the complex ways in which anthropomorphism and subjectivity can cloud our ability
to make reliable judgements concerning the welfare of companion animals, even in the face of seemingly obvious and overt indica-
tors of pain and suffering.
Keywords:animal welfare, anthropomorphism, companion animals, health, pets, subjectivity
Introduction
Detachment and objectivity are values that inform how
science, ideally, should be practiced. When attempting to
uncover truths about some aspect of the natural world, scien-
tists are expected to be aware of, and to attempt to eliminate,
personal biases, prejudices, and a priori commitments while
also detaching themselves emotionally from their subject
matter (Howard 1985). In practice, of course, true objectivity
in science is aspirational rather than fully achievable.
Scientists are only human and, like everyone else, may find
it difficult to remain detached and objective when they have
a strong personal interest in their object of study.
An illustration of the nature of this problem is provided by
Fraser et al (1997) who posed a hypothetical scenario
involving two dog owners who meet while walking their
dogs. One of these owners:
...had grown up in a small family that valued health,
safety and orderly, disciplined behaviour. The dog of
this owner received regular veterinary care, two meals a
day of low-fat dog food, and was walked on a leash.
The other owner had grown up in a large community
that valued conviviality, sharing of resources and close
contact with the natural world. This dog (the owner’s
third — the first two had been killed by cars) had burrs
in its coat, was fed generously but sporadically, and had
never worn a collar in its life. Each owner, judging
quality of life from very different viewpoints, felt sorry
for the other’s dog
The obvious point of the story is that each person tends to
view the quality of life of non-human animals through the
lens of his or her own subjective beliefs, attitudes, and
values, and that these inevitably colour perceptions of the
animals’ welfare. The conventional welfare scientist’s
response to this dilemma is to focus on the accurate and
objective measurement of things that are likely to be
relevant to the welfare of these two dogs — eg regularity
of meals, levels of physical restraint, risks of injury, and so
on — while also acknowledging that the ultimate determi-
nation of how well or poorly each dog is faring is going to
be largely subjective, since it will depend on the indi-
vidual, and potentially biased, perceptions and beliefs of
whoever is making the judgement (Mason & Mendl 1993;
Fraser 1995). If this is the case, however, it raises
important questions regarding the particular factors that
may interfere with our ability to recognise and prioritise
the things that actually matter to the animals, and hence
make sensible and effective recommendations to improve
their welfare. In Fraser et al’s (1997) example, the
emphasis is on differences in personal values — health,
safety and discipline versus laissez faire communal life
and contact with nature. The current discussion will focus
instead on the influence of past and current relationships
with animals, particularly companion animals, on our
ability to make objective assessments of their welfare.
Universities Federation for Animal Welfare Science in the Service of Animal Welfare

58 Serpell
Human-animal relationships as obstacles to
objectivity
The tendency to attribute ‘minds’ or mental states to others
(‘theory of mind’) is thought to have evolved as a social
adaptation that allows normal adult humans to use intro-
spection and self-knowledge as a model for understanding
and anticipating the feelings and behaviour of other humans
(Humphrey 1983; Mithen 1996). Since humans are so
intensely social, however, their desire for connection often
leads them to imbue non-human as well as human others
with human-like traits (Epley et al 2008; Waytz et al 2010).
This predisposition is generally termed ‘anthropomor-
phism’ and is usually defined as the tendency to attribute
human-like mental capacities, such as intentionality,
emotions, and cognition, to non-human agents and entities
(Epley et al 2008). Anthropomorphic attributions also affect
the moral status given to non-humans. Beings which are
believed to possess human-like minds are typically afforded
greater moral consideration and better treatment than those
deemed to have lesser mental capacities (Gray et al 2012).
Not surprisingly, the tendency to anthropomorphise is
particularly strong in relation to non-human animals
(henceforth ‘animals’), especially those that are phyloge-
netically close to, or which resemble, humans either phys-
ically or behaviourally (Burghardt & Herzog 1989; Plous
1993; Serpell 2004). Anthropomorphism also appears to
be amplified by social relationships and attachments with
particular animals, such as pets, especially if these attach-
ments develop early in life (Myers & Saunders 2002).
Numerous studies have demonstrated associations
between childhood pet ownership and the development of
positive attitudes and practices toward animals in
adulthood, including an increased likelihood of owning
companion animals in the future, more sympathetic views
on the treatment of both companion and non-companion
animals, greater willingness to support animal protection
causes, and a propensity to avoid eating certain animal-
based food products (Serpell 1981; Paul & Serpell 1993;
Paul 2000; Rothgerber & Mican 2014). Studies of veteri-
nary students further indicate that early affiliative relation-
ships with animals are strong predictors of later
professional orientations. For instance, veterinary students
who grow up with household dogs and cats are more likely
to gravitate toward careers in small animal practice, while
those who grow up with horses and ponies are more
strongly inclined toward equine practice (Serpell 2005). It
is also likely that anthropomorphism — in this case, the
ability to attribute human-like social motivations and
needs to companion animals — is what ultimately enables
these animals to provide their owners with the widely
reported psychological and physical benefits associated
with pet ownership (Serpell 2003; Epley et al 2008).
Their apparent ability to trigger ‘innate’ parental motiva-
tions, further suggests that companion animals, or at least
some companion animals, may act as supernormal stimuli.
The term ‘supernormal stimulus’ was first coined by the
Dutch ethologist and Nobel laureate, Niko Tinbergen, to
describe the tendency of animals (including humans) to
display open-ended preferences for biologically relevant
stimuli that are more extreme or exaggerated than would
occur in nature (Tinbergen 1951). For example, Tinbergen
observed that, when offered the choice between a normal
egg and an artificial, supersized one, nesting oystercatchers
(Haematopus ostralegus) will attempt to incubate the unnat-
urally large egg in preference to their own normally sized
one. Extrapolating from this original concept, various
authors subsequently concluded that the intrinsically
appealing, anthropomorphic or paedomorphic features of
things like teddy bears, Disney cartoon characters, and
many pets, possess supernormal stimulus properties that
appear ‘designed’ to elicit human parenting or nurturing
behaviour (Lorenz 1943; Gould 1979; Frank & Frank 1982;
Serpell 1996; Archer 1997; Chersini et al 2018). In some
cases, this idea has been used to support the theory that pet-
keeping is essentially a form of social parasitism (eg Archer
1997, 2011) in which the animals are seen as exploiting a
novel ecological niche provided by humans’ inflexible
parenting instincts. Others have proposed a less one-sided,
more mutualistic interpretation in which both species poten-
tially benefit from the relationship, though clearly not in the
same ways (Serpell 1996; Serpell & Paul 2011). For the
animals, the material and biological advantages of this asso-
ciation are relatively obvious. For the humans, a growing
body of evidence suggests that interactions with companion
animals stimulate fundamental attachment processes
mediated by the release of the bonding hormone, oxytocin,
which also appears to mitigate psychosocial stress (Julius
et al 2013; Serpell et al 2017).
Anthropomorphism and paedomorphism also have conse-
quences for animal welfare that may be either positive or
negative depending on the circumstances. On the positive
side, perceiving an animal to have a mind like one’s own
implies that it is capable of experiencing conscious
feelings and emotions, and that it is therefore worthy of
greater moral consideration (Bastian et al 2012; Gray et al
2012). Clearly, in this context, anthropomorphism has the
potential to arouse greater sensitivity to perceived welfare
problems in animals, thereby making it less likely that
these problems will be ignored or overlooked. Lockwood
(2005), for example, has noted that animal protection
supporters and activists in the USA are usually either
current or former pet owners. Similarly, members of the
pet-owning public have a tendency to react with moral
outrage toward activities such as dog- or cat-eating,
commercial breeding of pets, or the use of dogs and cats in
biomedical research, while typically accepting the similar
treatment of non-pet species (Serpell 2009). As previously
suggested, anthropomorphism may also serve as an
important motivator; encouraging people to try and help
animals by becoming ethical vegetarians, animal activists,
veterinarians, animal welfare scientists, and so on.
Whether or not this greater concern for animal well-being
among pet owners accurately and objectively reflects the
animals’ true quality of life, however, is somewhat ques-
tionable. Other animals clearly have different needs,
© 2019 Universities Federation for Animal Welfare

Assessment of welfare in companion animals 59
interests and cognitive capacities than humans, so using a
human or child-like model as a guide to their welfare is
likely to lead to at least some level of misunderstanding.
For example, anthropomorphic attributions imply that
animals, like humans, are capable of premeditated inten-
tional acts and that they can therefore, in principle, be held
responsible for those acts (Waytz et al 2010). An example of
the potential negative welfare consequences of this effect is
the widespread belief among dog owners that their pets are
capable of secondary emotions, such as guilt when they
steal food, raid the trash, or soil the carpet while the owner
is out of the house. Current evidence suggests that the
majority of owners interpret the dog’s ‘guilty look’ when
they return home as evidence that it is fully aware of its
transgression and therefore culpable. In reality, the results
of controlled experiments suggest that dogs respond with
characteristic guilty-looking behaviour when scolded by
their owners regardless of whether or not they are actually
guilty of any misdeed (Horowitz 2009). In such cases, close
relationships and anthropomorphic attributions lead owners
to subjectively overestimate the cognitive capacities of their
pets and to punish them accordingly.
When they predispose owners to prioritise the quantity of
an animal’s life over its quality, strong anthropomorphic
attachments can also give rise to severe and prolonged
discomfort and distress in companion animals. A growing
problem in small animal veterinary practice is the propen-
sity of owners to reject euthanasia for terminally ill and
suffering pets in favour of prolonging the animal’s life at
any cost (Sandøe et al 2016; Knesl et al 2017). While this
is partly a consequence of veterinary medical advances
and the increasing availability of previously inaccessible
treatment options, it also reflects the tendency of owners
to project essentially human notions of the value or
‘sanctity’ of life onto their pets rather than considering the
animal’s perspective. As Sandøe et al (2016) point out:
The human attachment to a companion animal can be
strong and highly emotionally charged, making it difficult
for some owners to be objective when it comes to making
decisions about their companion’s treatment
Objective evaluations of animal welfare become particu-
larly problematic when the traits that tend to enhance
people’s attachments to animals are themselves associated
with compromised health and welfare.
In sickness and in health: the dark side of
‘cuteness’
Some years ago, my son returned home with a three-week
old, tail-less kitten that he found abandoned in the street.
Fortunately, my veterinary contacts enabled us to obtain the
necessary advice and assistance to care for such a young and
helpless kitten, and after 2–3 weeks of bottle-feeding we were
able to wean him successfully onto solid food. During this
process, however, it became apparent that Henry (as we
called him) was not a normal kitten. In addition to carrying
the tail-less or ‘Manx’ gene, which is associated with spina
bifida, it turned out that he had also been infected in utero
with feline panleukopenia and was, as a result, suffering from
an untreatable neurological condition known as cerebellar
hypoplasia (CH), or ‘wobbly cat syndrome’. This meant that
he could scarcely walk more than a few steps without falling
over, displayed uncontrollable head tremors whenever he
tried to focus on anything, and had great difficulty using his
litter tray. At this point, we began to have some fraught
discussions about whether it would be kinder to euthanase
Henry rather than let him live with such profound disabilities.
The primary arguments in favour of the euthanasia option
were the potential for future suffering, and the fact that he
would never get any better or be able to engage in most of the
normal, species-typical activities that could be said to define
a cat’s nature or telos (sensu Rollin 1993). The argument
against hinged entirely on our subjective impression that, in
spite of his disability, Henry was not suffering anything
worse than periodic and transient inconvenience, and occa-
sional minor discomfort from his condition.
The second argument eventually prevailed, and Henry ulti-
mately developed into one of the most engaging and
rewarding cats we have ever lived with. His physical
disabilities, however, did not improve. If anything, they
became worse, at least with respect to consequences. His
lack of a tail and his long hind legs associated with the
Manx gene made his gait very unstable, and his larger size
meant that when he fell or blundered into chairs and table
legs the impact was more severe. He could not jump or
climb, and his attempts to chase toys or flies invariably
ended badly. Nevertheless, he appeared to be relatively
undaunted by his affliction. He learned to ‘fall’ through the
cat door to gain access to the back garden, he dominated
the family dog who was many times his size, embedded
his claws in people’s legs when he wanted to be picked up,
and never turned down an opportunity to play, despite the
often painful consequences. For a cat, he was also
intensely human-centered and sociable. He was unafraid
of strangers, seemed to enjoy being the centre of attention,
and appeared to like being picked up and cuddled. The
idea that we once seriously considered euthanasing him
began to seem callous and unethical — though this
opinion was not necessarily shared by our friends and
neighbours, some of whom clearly believed that it was
cruel and self-indulgent of us to keep Henry alive.
It turns out that our experience with Henry was not an
uncommon one. A Google or YouTube search for ‘wobbly
cat’ will usher one into a sort of parallel universe
containing countless affectionate videos and heart-
warming stories involving cats with CH. Without
exception, the owners of these animals describe them as
the most endearing cats they have ever owned, while
animal shelters and cat rescue groups who post videos of
CH cats online seem to be able to find homes for them
instantly while, at the same time, having great difficulty
adopting out large numbers of ostensibly normal, healthy
and homeless cats. So, what exactly is going on here? Why
are some people, including myself, apparently drawn to
these abnormal and physically compromised cats?
Animal Welfare 2019, 28: 57-66
doi: 10.7120/09627286.28.1.057

60 Serpell
One possible answer is that CH cats are, in fact, more affec-
tionate and friendly than normal cats, perhaps due to some
additional neurological change associated with cerebellar
hypoplasia or because of the effects of enhanced early
handling and socialisation. Certainly, in Henry’s case, he
received more attention in early life than a normal kitten
would have, and this may have increased his overall level of
sociability. One reason to question this as a general expla-
nation, however, is that the human affinity for sick and
disabled pets appears to extend far beyond wobbly cats (see,
for example, Segal 2011). In the animal-sheltering world, it
is widely recognised that it is often easier to find homes for
animals with medical problems or histories of abuse than it
is to adopt the normal ones. In some parts of the USA, for
instance, particular animal rescue groups specialise in
adopting and providing home hospice care for shelter cats
with end-stage renal disease. Since these cats would ordi-
narily be candidates for immediate euthanasia, this places
shelter veterinarians in the egregious position of being able
to re-home the terminally ill cats while having to euthanase
the healthy ones for lack of adoptive homes (B Watson,
personal communication 2018). I am not aware of any
studies of the motivations underlying this type of phenom-
enon, but it appears to be driven by individuals who find the
experience of caring for sick cats more rewarding than
caring for healthy ones, presumably because the former are,
by definition, more dependent and therefore more in need of
care. Nor is this propensity restricted to cats.
In a recent paper, Sandøe et al (2017) investigated the
human motivations underlying the continued (and in some
cases, growing) popularity of dog breeds that tend to suffer
from health and welfare problems due to the effects of
extreme conformation and/or genetically damaging
breeding practices. The survey was conducted on a random,
representative sample of Danish dog owners, and included
the owners of three affected breeds (Cavalier King Charles
spaniel, Chihuahua, and French bulldog) with extreme
phenotypes and/or genetic histories associated with health
problems, in addition to one unaffected breed (Cairn terrier)
that is similar in size but otherwise average in terms of its
health. A number of interesting findings emerged from this
study, but two are particularly relevant to the present discus-
sion. First, owners of the three affected breeds were more
attached to their dogs than were the owners of the healthier
Cairn terriers. Second, the perceived quality of owners’ rela-
tionships with these dogs was marginally positively associ-
ated (P< 0.07) with the number of frequently occurring
health problems they experienced. As with the CH cats, it
appears that these Danish dog owners showed a tendency to
prefer the less healthy dogs with the more extreme pheno-
types to the relatively normal and healthier ones.
Such preferences appear to apply to behavioural as well as
physical problems. In a series of studies, McMillan and
colleagues examined the behavioural characteristics of dogs
that had been rescued and re-homed from various chal-
lenging life situations: eg former breeder dogs from so-
called ‘puppy mills’ (McMillan et al 2011), dogs rescued
from ‘hoarding’ situations (McMillan et al 2016), and dogs
that had been victims of serious physical abuse (McMillan
et al 2015). In every case, these dogs exhibited a signifi-
cantly higher prevalence of certain behavioural problems,
particularly those associated with fear and anxiety, than
matched comparison samples of ‘normal’ pet dogs. In
several cases, these differences in behaviour were substan-
tial. Puppy mill ex-breeder dogs, for instance, were between
5 and 7 times more likely to display fear in response to
strangers, unfamiliar dogs, and non-social stimuli than were
the current pets (McMillan et al 2011). Yet, despite these
signs of chronic fear and anxiety, their owners’ evaluations
of their relationships with these dogs were overwhelmingly
positive. When McMillan questioned them regarding their
level of satisfaction with the dogs, and whether they would
consider adopting another dog of the same type in the
future, their responses to the first question were between
86 and 98% ‘extremely satisfied’ and only 0–2% ‘not
satisfied’, and when asked whether they would consider
adopting the same type of dog in the future, 93–100%
agreed that they would (McMillan 2014, 2016) (see
Figure 1). In other words, the experience of caring for these
often severely emotionally disabled dogs, far from being a
deterrent to future ownership, was apparently an added
incentive, at least for these individuals.
Evidence of the prevalence of fear- and anxiety-related
behaviour problems in the pet dog population suggests that
such incentives may be relatively widespread. The Canine
Behavioral Assessment and Research Questionnaire (C-
BARQ) database at the University of Pennsylvania has
been collecting standardised behavioural evaluations of pet
dogs from their owners via an online portal since 2005 and
has now amassed more than 50,000 such assessments. In
addition to addressing a variety of other aspects of
behaviour, the C-BARQ includes four separate fear-related
subscales — fear of strangers, fear of other dogs, fear of
non-social stimuli, and touch sensitivity — with scores
ranging from zero (absence of fearful responses) to 4
(extreme fearfulness). The population distribution of scores
for these different subscales tend to be skewed toward zero,
and yet between 14 and 18% of owners report that their
dogs display fearfulness in the moderate to extreme range
(scores of 2–4) in one or more of these contexts. More to
the point, many owners do not appear to regard their dogs’
extreme fear as being a problem. Among the background
information owners are asked to provide before completing
the C-BARQ is the question: ‘Are you currently experi-
encing any problems with your dog’s behaviour or
temperament?’ The offered response options are ‘no
problems’, ‘only minor problems’, ‘moderate problems’
and ‘serious problems.’ As might be expected, owners’
overall reported levels of perceived problems tend increase
in proportion to the severity of their dogs’ fear-related
behaviour (see Figure 2). However, it is also clear that there
are large numbers of outliers — large numbers of dogs with
moderate to extreme fearfulness scores on the C-BARQ
whose owners report experiencing ‘only minor problems’
or ‘no problems’ at all with their behaviour.
© 2019 Universities Federation for Animal Welfare

Assessment of welfare in companion animals 61
Animal Welfare 2019, 28: 57-66
doi: 10.7120/09627286.28.1.057
Figure 1
Willingness of adopters of dogs with extreme anxiety/fearful behaviour to consider re-adopting dogs of the same type again (F McMillan,
unpublished survey data, reproduced with permission).
Relationship between levels of fear/anxiety in pet dogs, as measured by the C-BARQ, and owner-reported experience of behavioural
problems (Boxplots showing medians, 10th, 25th, 75th and 90th percentiles and outliers).
Figure 2

62 Serpell
This result is reminiscent of those obtained in another
study of dogs suffering from brachycephalic obstructive
airway syndrome (BOAS). In this case, owners of
285 dogs belonging to various brachycephalic breeds were
asked whether their dog currently had, or had a history of,
breathing problems. Separately, the dogs were then
diagnosed as ‘affected’ or ‘unaffected’ with BOAS using
various medical criteria including the owners’ reports of
respiratory difficulties and noisy breathing in four
different standardised situations. More than half (58%) of
the owners of the affected dogs reported that their dog did
not have any breathing problems (Packer et al 2012). The
authors conclude that the disparity between the dogs’
severe clinical symptoms, and their owners’ perceptions of
no breathing problems would be likely to lead to these
animals suffering indefinitely due to lack of veterinary
treatment. Furthermore, in a recent, separate survey of
2,168 owners of brachycephalic dogs, Packer et al (2018)
report that many owners, when asked what they would
recommend most about their breed, freely admitted to
liking their dogs’ increased level of dependence on them
due to health and conformation problems.
Discussion
Just as modern dairy cows or broiler chickens are the
products of artificial selection for high milk yield and rapid
growth, respectively (Duncan 2001; Rushen 2001), so
companion animals — or at least some companion
animals — increasingly are products of selection for traits
that make them more emotionally appealing to people. For
many humans, that appeal lies primarily in their ability to
mimic and in some cases exaggerate the infantile cues that
stimulate parental nurturing responses and behaviour. The
characteristic physical features of such animals — their
small body size, short limbs relative to body size, soft skin
and fur, large eyes relative to head size, domed foreheads,
flattened muzzles, drooping ears, and so on — are perceived
as ‘cute’ and tend to elicit care-giving motivations from
people (Serpell 1996, 2003; Archer & Monton 2011; Little
2012; Golle et al 2013; Lehmann et al 2013; Waller et al
2013; Hecht & Horowitz 2015; Chersini et al 2018).
Similarly, chronic health problems, such as CH, and
behaviour problems, such as fearfulness and anxiety, that
tend to increase these animals’ perceived vulnerability and
dependency also appear to enhance their desirability as
objects of human care-giving and attachment. Thus, it could
be argued that humans have selected unconsciously for
small, anxious, needy, unhealthy and vulnerable companion
animals — animals with inherently compromised
welfare — because these are precisely the traits that best
satisfy their desire for things to nurture and parent.
Furthermore, this process appears to be ongoing. The evolu-
tionary trends in many dog and cat breeds seem to be toward
increasingly accentuated anthropomorphic and paedomor-
phic anatomical and behavioural features, despite — and to
some extent because of — their association with compro-
mised mental and physical health (see Figure 3; Serpell
2003). Thus, the bulbous, protruding eyes and febrile
anxiety of the Chihuahua, and the hopelessly wobbly gait of
the cat with cerebellar hypoplasia, are sources of attraction
and endearment to many pet owners precisely because of
their ability to evoke care-giving responses and the resulting
cascade of neurophysiological rewards associated with
these kinds of human-animal interactions.
These psychological processes are also likely to produce
parallel effects on pet owners’ moral judgements. Causing
harm to others for personal gain is one of the most widely
held moral prohibitions in human cultures throughout the
world (Graham et al 2009; Gray et al 2012). Since human
relations with animals commonly involve harming them,
either deliberately or inadvertently, for self-interested
reasons, humans have developed a variety of psychological
techniques to reduce the moral discomfort that arises
inevitably from this. Probably the most widespread of these
techniques has been variously termed ‘moral disengage-
ment’ (Bandura 1999; Vollum et al 2004) or ‘dissonance
avoidance’ (Bastian et al 2012; Bastian & Loughnan 2017)
and it typically includes such things as dehumanising
animals (the opposite of anthropomorphism), and avoiding
social engagement with them that might otherwise lead to
anthropomorphic attributions and an increase in moral
concern for their welfare (Serpell 1996, 1999). This
phenomenon has been demonstrated in a variety of
contexts. For example, Bastian et al (2012) showed exper-
imentally that human subjects are less likely to ascribe
higher mental capacities to animals they consider appro-
priate to eat and are also more inclined to deny minds to
food animals when reminded of the link between meat and
animal suffering. The ability to identify and respond to
suffering in animals may be similarly influenced by self-
interested motives. In a study of dairy farm managers, for
instance, Šárová et al (2011) found that they consistently
underestimated the actual prevalence of lameness in their
cows, presumably because acknowledging the true preva-
lence would cause moral conflict as well as entailing poten-
tially costly remedial interventions.
This type of moral disengagement or dissonance avoidance
is not readily available to pet owners. The ostensible benefit
and purpose of pet-keeping is to provide owners with a
source of non-human social and emotional support, and this
function is partly predicated on owners attributing human-
like or child-like feelings and cognitions to their pets
(Serpell 2003). However, since this function sometimes
causes or perpetuates animal suffering, albeit indirectly, pet
owners must resort to other methods to alleviate moral
responsibility for these outcomes. Dehumanising or de-
mentalising the pet would clearly defeat the whole object of
the exercise, so other options must be found. In the cases
described in this paper, these seem to involve either denying
or minimising the welfare problems that currently exist in
companion animals, or categorising them as ‘normal’ for the
particular breeds that are affected. The tendency of pet
owners to cast themselves as benevolent animal lovers may
also help to reduce dissonance by feeding the perception that
they are really acting in the animal’s best interests by caring
© 2019 Universities Federation for Animal Welfare

Assessment of welfare in companion animals 63
for it and keeping it ‘happy’ (Bastian et al 2012). For these
reasons, asking current or former owners of such animals to
provide reliable and objective evaluations of their welfare is
likely to be as fruitless as asking dairy farmers to make
comparable assessments of their cows. More to the point,
pet-keeping is considerably more common than dairy
farming, and many if not the majority of animal advocates,
veterinarians and animal welfare scientists are probably
current or former pet owners with histories of strong
emotional attachments to companion animals (Paul &
Serpell 1993; Serpell 2005). All of which raises challenging
questions regarding the extent to which pet owners in
general, and animal welfare ‘experts’ in particular, can
achieve sufficient psychological distance from these animals
to allow truly objective assessments of their welfare.
While it helps to explain the apparent paradox of people
claiming to be animal lovers while simultaneously helping to
perpetuate severe welfare problems in the objects of their
affection, the proposed theory is not without weaknesses.
Not all companion animal breeds, for example, have been
equally affected by anthropomorphic selection pressures,
and not all pet owners are necessarily attracted to the most
anthropomorphic or infantilised breeds with the most
extreme welfare problems. Human preferences for animal
companions are clearly complicated and not easily explained
by any single all-encompassing hypothesis. Hopefully,
future studies of the psychological origins of such prefer-
ences will help to clarify the motivations underlying human
predilections for particular pets, as well as providing
direction on how to improve companion animal welfare by
modifying people’s pet-related attitudes and behaviour.
Animal Welfare 2019, 28: 57-66
doi: 10.7120/09627286.28.1.057
Figure 3
Morphological trends in the English bulldog breed illustrating the gradual exaggeration of anthropomorphic/paedomorphic features over
historical time: A) 1859; B) 1889; C) 1903; and D) 2013.

64 Serpell
Animal welfare implications
Accurately evaluating an animal’s quality of life demands a
degree of scientific objectivity. It is therefore important to
understand the factors that may limit or impede the
objective assessment of welfare. People’s attitudes to, and
relationships with, companion animals are often highly
anthropomorphic and subjective. Attributing human-like
minds and motives to these animals is probably essential to
their function as social support providers, and this in turn
has driven selection for morphological and behavioural
traits in some companion animal species/breeds that facili-
tate anthropomorphic and paedomorphic attributions. Many
of these traits are also inherently detrimental to the animals’
health and welfare. If, as suggested, the rewards of pet
ownership are partly conditional on these traits, this is likely
to create psychological resistance to addressing some of the
most pressing welfare problems that currently exist, and
which continue to be perpetuated, in these kinds of animals.
Overcoming this resistance in order to effect long-term
improvements in the health and welfare of companion
animals will require novel programmes and policies to
increase awareness among prospective pet owners that the
very attributes they find so attractive and appealing are also
those likely to cause their animals lifetimes of distress and
discomfort. It will also require pressure on the animal
breeding community to identify and de-accentuate the
various traits that contribute significantly to conformation-
related health and welfare problems. If such efforts fail, the
welfare problems of companion animals are likely to
increase in frequency and severity until they eventually
become self-limiting — ie, when the financial and
emotional costs to owners of maintaining the health of their
pets exceed the psychological rewards derived from
keeping them. Efforts to arrest this process before it reaches
such an extreme and harmful endpoint would be beneficial
to the welfare of both pets and their owners.
Conclusion
McMullin (1982) wrote that emotive values such as attrac-
tion, feelings and emotions are alien to the work of natural
science:
There is no reason to think that human emotionality is a
trustworthy guide to the structures of the natural world.
Indeed, there is every reason, historically speaking, to
view emotive values, as Bacon did, as potentially
distortive ‘idols’, projecting in anthropomorphic
fashion the pattern of human wants, desires, and
emotions on a world where they have no place
A major effect of human selection on the evolution of
companion animals has been to accentuate those morpho-
logical and behavioural characteristics that elicit strong
emotive values. In this sense, it could be said that many
modern companion animals are quite literally becoming
anthropomorphic or paedomorphic projections of ‘human
wants, desires and emotions.’ In consequence, these types of
animals present uniquely challenging obstacles to the
objective assessment of their quality of life. Somewhat iron-
ically, while pet-keeping may have provided many animal
welfare scientists (myself included) with the original
impetus to pursue their chosen careers, it may also obstruct
their ability to evaluate the welfare of these animals objec-
tively. Future studies of companion animal welfare, and
programmes and policies designed to change public
attitudes and behaviour towards these animals, will need to
be particularly alert to the potential difficulties created by
these sorts of inherent biases.
References
Archer J 1997 Why do people love their pets? Evolution and
Human Behaviour 18: 237-259. https://doi.org/10.1016/S0162-
3095(99)80001-4
Archer J 2011 Pet keeping: A case study in maladaptive behavior. In:
Salmon C and Shackelford TK (eds) Oxford Handbook of Evolutionary
Family Psychology pp 281-296. Oxford University Press: Oxford, UK.
https://doi.org/10.1093/oxfordhb/9780195396690.013.0016
Archer J and Monton S 2011 Preferences for infant facial fea-
tures in pet dogs and cats. Ethology: International Journal of
Behavioral Biology 111(3): 217-226. https://doi.org/10.1111/j.1439-
0310.2010.01863.x
Bandura A 1999 Moral disengagement in the perpetration of
inhumanities. Personality and Social Psychology Review 3: 193-209.
https://doi.org/10.1207/s15327957pspr0303_3
Bastian B and Loughnan S 2017 Resolving the meat-paradox:
A motivational account of morally troublesome behavior and its
maintenance. Personality and Social Psychology Review 21: 278-291.
https://doi.org/10.1177/1088868316647562
Bastian B, Loughnan S, Haslam N and Radke HRM 2012
Don’t mind meat: The denial of mind to animals used for human
consumption. Personality & Social Psychology Bulletin 38: 247.
https://doi.org/10.1177/0146167211424291
Burghardt GM and Herzog HA 1989 Animals, evolution, and
ethics. In: Hoage JH (ed) Perceptions of Animals in American Culture
pp 129-151. Smithsonian Institution Press: Washington, DC, USA
Chersini N, Hall NJ and Wynne C 2018. Dog pups’ attractive-
ness to humans peaks at weaning age. Anthrozoös 31: 309-318.
https://doi.org/10.1080/08927936.2018.1455454
Duncan IJH 2001 Animal welfare issues in the poultry industry: Is
there a lesson to be learned. Journal of Applied Animal Welfare Science
4: 207-221. https://doi.org/10.1207/S15327604JAWS0403_04
Epley N, Waytz A, Akalis S and Cacioppo J T 2008 When we
need a human: Motivational determinants of anthropomorphism. Social
Cognition 26: 143-155. https://doi.org/10.1521/soco.2008.26.2.143
Frank H and Frank MG 1982 On the effects of domestication on
canine social development and behaviour. Applied Animal Ethology 8:
507-522. https://doi.org/10.1016/0304-3762(82)90215-2
Fraser D 1995 Science, values and animal welfare: Exploring the
‘inextricable connection’. Animal Welfare 4: 103-117
Fraser D, Weary DM, Pajor EA and Milligan BN 1997 A
scientific conception of animal welfare that reflects ethical con-
cerns. Animal Welfare 6: 187-205
Golle J, Lisibach S, Mast FW and Lobmaier JS 2013 Sweet
puppies and cute babies: Perceptual adaptation to babyfacedness
transfers across species. PLoS One 8(3): e58248.
https://doi.org/10.1371/journal.pone.0058248
Gould SJ 1979 Mickey Mouse meets Konrad Lorenz. Natural
History 88: 30-36
© 2019 Universities Federation for Animal Welfare

Assessment of welfare in companion animals 65
Graham J, Haidt J and Nosek BA 2009. Liberals and conser-
vatives rely on different sets of moral foundations. Journal of
Personality and Social Psychology 96: 1029-1046.
https://doi.org/10.1037/a0015141
Gray K, Young L and Waytz A 2012 Mind perception is the
essence of morality. Psychological Inquiry 23: 101-124.
https://doi.org/10.1080/1047840X.2012.651387
Hecht J and Horowitz A 2015 Seeing dogs: Human prefer-
ences for dog physical attributes. Anthrozoös 28(1): 153-163.
https://doi.org/10.2752/089279315X14129350722217
Horowitz A 2009 Disambiguating the ‘guilty look’: Salient
prompts to a familiar dog behaviour. Behavioural Processes 81: 447-
452. https://doi.org/10.1016/j.beproc.2009.03.014
Howard GS 1985 The role of values in the science of psycholo-
gy. American Psychologist 40: 255-265. https://doi.org/1
0.1037/0003-066X.40.3.255
Humphrey N 1983 Consciousness Regained. Oxford University
Press: Oxford, UK
Julius H, Beetz A, Kotrschal K, Turner D and Uvnäs-
Moberg K 2013 Attachment to Pets: An Integrative View of Human-
Animal Relationships with Implications for Therapeutic Practice.
Hogrefe: Göttingen, Germany
Knesl O, Hart BL, Fine AH, Cooper L, Patterson-Kane E,
Houlihan KE and Anthony R 2017 Veterinarians and humane
endings: When is it the right time to euthanize a companion ani-
mal? Frontiers in Veterinary Science 4: 45. https://doi.org/10.3389
/fvets.2017.00045
Lehmann V, Huis i’nt Veld EM and Vingerhoets AJ 2013
The human and animal baby schema effect: Correlates of individ-
ual differences. Behavioral Processes 94: 99-108. https://doi.org/
10.1016/j.beproc.2013.01.001
Little AC 2012 Manipulation of infant-like traits affects perceived
cuteness of infant, adult and cat faces. Ethology 118: 775-782.
https://doi.org/10.1111/j.1439-0310.2012.02068.x
Lockwood R 2005 Tracking the ‘State of the Animals’:
Challenges and opportunities in assessing change. In: Salem DJ and
Rowan AN (eds) The State of the Animals III, 2005 pp 1-14.
Humane Society Press: Washington, DC, USA
Lorenz K 1943 Die angeborenen formen möglicher erfahrung.
Zeitschrift für Tierpsychologie 5: 235-409. [Title translation: The
innate forms of potential experience]. https://doi.org/10.1111/
j.1439-0310.1943.tb00655.x
Mason GJ and Mendl M 1993 Why is there no simple way of
measuring animal welfare? Animal Welfare 2: 301-320
McMillan FD 2014 Understanding and Caring for Rescued Puppy
Mill Dogs. Best Friends Animal Society: Kanab, UT, USA.
http://815678169699-bfas-files.s3-us-west-2.amazonaws.com/
Understanding-and-Caring-for-Puppy-Mill-Dogs.pdf
McMillan FD 2016 Understanding and Caring for Rescued Hoarded
Dogs. Best Friends Animal Society: Kanab, UT, USA.
http://815678169699-bfas-files.s3.amazonaws.com
/i n line _ image s/res ource s /639 2 -Bran ding- H oard e d-Do g s-
Manual.pdf
McMillan FD, Duffy DL and Serpell JA 2011 Mental health of
dogs formerly used as ‘breeding stock’ in commercial breeding
establishments. Applied Animal Behaviour Science 135: 86-94.
https://doi.org/10.1016/j.applanim.2011.09.006
McMillan FD, Duffy DL, Zawistowski SL and Serpell JA
2015 Behavioral and psychological characteristics of canine vic-
tims of abuse. Journal of Applied Animal Welfare Science 18: 92-111.
https://doi.org/10.1080/10888705.2014.962230
McMillan FD, Vanderstichel R, Stryhn H, Yu J and Serpell
JA 2016 Differences in behavioral characteristics between dogs
removed from hoarding situations and a convenience sample of
pet dogs. Applied Animal Behaviour Science 178: 69-79.
https://doi.org/10.1016/j.applanim.2016.02.006
McMullin E 1982 Values in science. Proceedings of the Biennial
Meeting of the Philosophy of Science Association 2(2): 3-28.
https://doi.org/10.1086/psaprocbienmeetp.1982.2.192409
Mithen S 1996 The Prehistory of the Mind. Thames & Hudson:
London, UK
Myers OE and Saunders CD 2002 Animals as links toward
developing caring relationships with the natural world. In: Kahn
PH and Kellert SR (eds) Children and Nature: Psychological,
Sociocultural and Evolutionary Investigations pp 152-178. MIT Press:
Cambridge, MA, USA
Packer RMA, Fletcher F and Farnworth MJ 2018 ‘Could not
live without one.’ Factors driving repeated purchase and recom-
mendation of brachycephalic dog breeds (abstract). Proceedings of
UFAW Animal Welfare Conference. 28th June 2018, Newcastle-
upon-Tyne, UK. https://www.ufaw.org.uk/downloads/programme-
booklet-for-ufaw-newcastle-conference-2018—-final.pdf
Packer RMA, Hendricks A and Burn CC 2012 Do dog own-
ers perceive clinical signs related to conformational inherited dis-
orders as ‘normal’ for the breed? A potential constraint to
improving canine welfare. Animal Welfare 21: 81-93.
https://doi.org/10.7120/096272812X13345905673809
Paul ES 2000 Empathy with animals and with humans: Are they linked?
Anthrozoös 13: 194-202. https://doi.org/10.2752/089279300786999699
Paul ES and Serpell JA 1993 Childhood pet keeping and
humane attitudes in young adulthood. Animal Welfare 2: 321-237
Plous S 1993 The role of animals in human society. Journal of
Social Issues 49: 1-9. https://doi.org/10.1111/j.1540-
4560.1993.tb00906.x
Rollin BE 1993 Animal welfare, science and value. Journal of
Agricultural and Environmental Ethics 6(2): 44-50
Rothgerber H and Mican F 2014 Childhood pet ownership,
attachment to pets, and subsequent meat avoidance. The mediat-
ing role of empathy toward animals. Appetite 79: 11-17.
https://doi.org/10.1016/j.appet.2014.03.032
Rushen J 2001 Assessing the welfare of dairy cattle. Journal of
Applied Animal Welfare Science 4: 223-234. https://doi.org/10.1207
/S15327604JAWS0403_05
Sandøe P, Corr S and Palmer C 2016 Companion Animal
Ethics. Wiley Blackwell: Oxford, UK
Sandøe P, Forkman B, Kondrup SV, Lund TB, Meyer I,
Bennett P, Prochowsky HF and Serpell JA 2017 Why do
people buy dogs with potential welfare problems related to
extreme conformation and inherited disease? A representative
study of Danish owners of four small dog breeds. PLoS One 12(2):
e0172091. https://doi.org/10.1371/journal.pone.0172091
Šárová R, Stěhulová I, Kratinová P, Firla P and Špinka M
2011 Farm managers underestimate lameness prevalence in
Czech dairy herds. Animal Welfare 20: 201-204
Animal Welfare 2019, 28: 57-66
doi: 10.7120/09627286.28.1.057

66 Serpell
Segal J 2011 Feel Better Little Buddy: Animals with Casts. Chronicle
Books: San Francisco, CA, USA
Serpell JA 1981 Childhood pets and their influence on adults’
attitudes. Psychological Reports 49: 651-654.
https://doi.org/10.2466/pr0.1981.49.2.651
Serpell JA 1996 In the Company of Animals: A Study of Human-animal
Relationships, Second Edition. Cambridge University Press: Cambridge, UK
Serpell JA 1999 Sheep in wolves’ clothing: Attitudes to animals
among farmers and scientists. In: Dolins FL (ed) Attitudes to
Animals: Views in Animal Welfare pp 26-33. Cambridge University
Press: Cambridge, UK. https://doi.org/10.1017/
CBO9780511608476.003
Serpell JA 2003 Anthropomorphism and anthropomorphic
selection—beyond the ‘cute response’. Society & Animals 11: 83-
100. https://doi.org/10.1163/156853003321618864
Serpell JA 2004 Factors influencing human attitudes to animals
and their welfare. Animal Welfare 13(S): 145-151
Serpell JA 2005 Factors influencing veterinary students’ career
choices, and attitudes to animals. Journal of Veterinary Medical
Education 32: 491-496. https://doi.org/10.3138/jvme.32.4.491
Serpell JA 2009. Having our dogs and eating them too: Why ani-
mals are a social issue. Journal of Social Issues 65: 633-644.
https://doi.org/10.1111/j.1540-4560.2009.01617.x
Serpell JA, McCune S, Gee N and Griffin J 2017 Challenges
to research on animal-assisted interventions. Applied
Developmental Science 21: 223-233. https://doi.org/
10.1080/10888691.2016.1262775
Serpell JA and Paul ES 2011 Pets in the family: An evolution-
ary perspective. In: Salmon C and Shackelford TK (eds) Oxford
Handbook of Evolutionary Family Psychology pp 297-309. Oxford
University Press: Oxford, UK. https://doi.org/10.1093/oxford-
hb/9780195396690.013.0017
Tinbergen N 1951 The Study of Instinct. Oxford University Press:
Oxford, UK
Vollum S, Buffington-Vollum J and Longmir DR 2004
Moral disengagement and attitudes about violence toward nonhu-
man animals. Society and Animals 12: 209-236. https://doi.org/
10.1163/1568530042880668
Waller BM, Pierce K, Caeiro CC, Scheider L, Burrows
AM, McCune S and Kaminski J 2013 Paedomorphic facial
expressions give dogs a selective advantage. PLoS One 8(12):
e82686. https://doi.org/10.1371/journal.pone.0082686
Waytz A, Gray K, Epley N and Wegner DM 2010 Causes
and consequences of mind perception. Trends in Cognitive Sciences
14: 383-388. https://doi.org/10.1016/j.tics.2010.05.006
© 2019 Universities Federation for Animal Welfare